"On Plato's Sophist," Review of Metaphysics 46, No. 4 (June 1993): 747-780. Reprinted in The Argument of the Action, 2000.
Once the stranger takes over the discussion at the beginning of the Sophist, and agrees to discuss the sophist, the statesman, and the philosopher, it is hard to remember that Socrates once more, even after he had left Theaetetus completely barren, at least temporarily, and had encountered a resistance on Theodorus’s part to his further participation in any argument which the interval of a single day could not, it seems, have overcome. The Stranger’s intrusion thus makes us fail to notice that the only possible itnerest Socrates could have had in the same group would have been in young Socrates, about whom he knows only that he developed with Theaetetus a way of classifying two kinds of numbers: those with integral square or cube roots and thought without. If Socrates had engaged young Socrates in a discussion, and informed Euclides about it in the same way as he had reported his discussion with Theodorus and Theaetetus, we know that Euclides could not have transposed Socrates’ report into direct discourse and omitted Socrates’ “I said” and “He said.” A transposition of the kind Euclides practices in the Theaetetus would have led to the indiscernability of the two Socrateses, since each would have addressed the other as Socrates, and there is no reason to believe that the wise answers would have consistently belonged to only one of them.